Skip to main content
Campus Calendar

Seminar: Developing Trustworthy AI for Weather and Climate

  • To
  • Atlantic Building
Amy McGovern

Abstract:

As our climate is changing, the need to improve our understanding and prediction of a wide variety of high-impact weather phenomena has become increasingly urgent. The use of artificial intelligence (AI) and machine learning (ML) have grown rapidly for both weather and climate applications in the research community, but the uptake of AI/ML methods for operational use has been limited. The NSF AI Institute for Research on Trustworthy AI in Weather, Climate, and Coastal Oceanography (AI2ES) was formed in 2020 to create trustworthy AI for a variety of weather, climate, and coastal oceanography applications. In this talk, I give an overview of the convergent research we are doing in AI2ES to understand the nature of trust in AI and to create trustworthy AI. AI2ES is advancing AI/ML techniques in physics-based AI, explainable AI (XAI) and robust AI. We are developing guidelines to ensure that our AI is developed and deployed in an ethical and responsible manner. I will give deeper examples from work within my research group on autonomous detection of frontal boundaries and convective hazards and will discuss the concept of trust in AI from its foundations to how we have worked with operational forecasters to enhance trustworthy AI within meteorology. Finally, I will propose the development of a joint center for AI for weather and climate, whose goals are to work across agencies to develop foundational new AI/ML methods and apply them operationally to improve our resiliency to climate change and its related high-impact weather.

Location

Atlantic Building

Atlantic Building Room 2400

Contact

Department of Atmospheric & Oceanic Science

For disability accommodations, please contact Walter Tribett at wtribett@umd.edu

Hosted By

Event Tags

Audience